Chapter 3. Current trends in decentralisation

This chapter highlights that there has been a path towards decentralisation in a majority of OECD countries over the past decades. The chapter also underlines that deecentralisation trends around the world have often gone hand in hand with an upscale in subnational governance through amalgamations, inter-municipal cooperation, metropolitan governance, and the strengthening of regions. Municipal fragmentation has driven policies encouraging or imposing mergers or cooperation. In parallel, there has also been an increase in asymmetric decentralisation, i.e. the fact that governments at the same subnational government level have different political, administrative or fiscal powers. While asymmetric decentralisation appears more “natural” in federal countries, it is increasing in unitary countries. Finally, the chapter shows that decentralisation is too often understood as a simple increase in the power of local governments. The reality is much more complex, as most responsibilities are shared across levels of government. Decentralisation is about reconfiguring relationships between the central government and subnational governments towards a greater cooperation and a strategic role for national/federal governments.


The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

This chapter focuses on trends in decentralisation among OECD countries and around the world. Over the past 70 years, the overall trend has been in favour of greater decentralisation. It is sometimes called the “silent revolution” (Ivanyna and Shah, 2014[1]). Indicators that measure the authority of administrative regions (Regional Authority Index, RAI) show that all world regions are experiencing an increase in the RAI: western countries (mostly European) since the 1960s/1970s; Asia and Pacific countries since the 1980s; and Latin American countries since the 1980s, though to a lesser extent (Hooghe et al., 2016[2]). In Europe, the Local Autonomy Index (LAI) shows an increase of local autonomy between 1990 and 2005, especially in the Central and East European countries (Ladner, Keuffer and Baldersheim, 2016[3]).

Several OECD countries, both federal and unitary countries, already have a solid, long-established tradition of subnational self-government. The decentralisation trend has intensified continuously over the last few decades, in particular in unitary countries. Some other countries have experienced a back and forth fluctuation between decentralisation and recentralisation.

The global crisis in 2008-09 led to a certain recentralisation in some countries, with an increase in central government grants to support subnational governments. This was only a temporary trend, however, as it was followed by important cuts in central government grants after 2010-11 in most countries. Many countries, including almost one-third of OECD members, introduced fiscal rules in order to control subnational expenditures in the wake of the crisis. It is also possible that this enhanced control calls for greater intergovernmental co-ordination, which boosts the bargaining power of subnational jurisdictions to influence national policymaking (de Mello, 2018[4])

Several complementary trends in decentralisation stand out and should be understood in the context of the mutual dependence that exists among levels of government: i) increased subnational spending and revenues; ii) reinforced local autonomy (municipal authority); iii) an upscale in subnational governance through inter-municipal co-operation, metropolitan governance and the strengthening of regions; and iv) increased asymmetric decentralisation, i.e. the fact that responsibilities may vary across subnational governments.

Increased subnational spending and revenues in OECD countries

An increase in subnational spending

In two thirds of OECD countries, decentralisation processes have resulted in an increase of economic importance of subnational government, measured both as a spending share of GDP and share of total public spending between 1995 and 2016 (Figure 3.1). The highest increases over the last 21 years (1995-2016) were seen in Poland, Germany, Denmark, Belgium, Sweden and Spain. Spain, in particular, experienced the highest increase. In Spain, the education and health decentralisation process in the 2002 and 2005 resulted in an increase of 13 percentage points of the share of subnational government in public expenditure and of 5 percentage points in the subnational expenditure contribution to the GDP between 1995 and 2016. In Belgium, the significant increase of the weight in GDP (6 percentage points) and in public expenditure (11 points) is explained by the implementation of the 6th state reform of 2011 (in effect since 2014), which has devolved new responsibilities from the federal government to the regions and communities. In Sweden, there has been a continuous growth of subnational spending over the period. Swedish subnational government handles not just the “pure local public services” but also many of the redistribution functions of a welfare state, such as education and healthcare (OECD, 2017[5]). In Poland, the increase in subnational expenditure started prior to 1995, after the adoption of the 1990 Act on Municipality re-establishing municipal autonomy and gave them large responsibilities. There was a second peak in 1999 when the regional and intermediate levels were created to take care of several responsibilities such as education, health, social protection, transport and regional economic development, and then in 2009, when new tasks were transferred again to the regions.

In other OECD countries, the increase in subnational expenditure over the period may also be explained by the decentralisation process, such as in Germany, Denmark, Canada, etc. (see Annex C) although additional factors can explain this increase e.g. greater social needs, in particular with the crisis, an ageing population, increasing environmental and security norms and quality standards as well as the costs of services.

By contrast, several OECD countries have experienced a reduction in the share of subnational government (SNG] expenditure in GDP and/or in general government over the past two decades. In Hungary, there was the recentralisation reform, started in 2011-12 with the constitutional reform and the Local Government Law which led the central government to reorganise the Hungarian local government sector and take over many functions exercised previously by municipalities and counties. As a result, Hungary went from being quite decentralised compared to other OECD countries to being among the most centralised. In 2010, subnational expenditure amounted to 25% of public expenditure, i.e. 12% of GDP to respectively 13% and 6% in 2017, a reduction by half in both cases (OECD, 2018[6]).

This decrease in subnational spending may also be the result of other trends, in particular since the global crisis that put subnational budgets under strong pressures. Therefore, in several countries, a decrease in subnational spending can be explained more by the effects of fiscal consolidation measures (spending cuts, savings programmes) or public management reforms aimed at seeking effectiveness and cost-efficiency (pooling of services and shared services agreements, performance assessments, public staff reforms, assets management, etc.) than a recentralisation process per se.

The case of Ireland is in between. A strong decrease in subnational spending in 2005, 2010 and in 2014 resulted both from recentralisation trends (e.g. water services and some other tasks in 2014), from Better Local Government reforms, including the 2014 Local Government Reform Act which merged 114 local councils into 31 local governments and abolished the previous 8 regional authorities, and from the 2008 recession. The crisis significantly reduced local income and necessitated spending and staffing cuts imposed by central government (Quinn, 2015[7]).

In the Netherlands, the decline in subnational expenditure was observed only in 1995-96 (subnational spending dropped from 40% of public spending to 33%, and from 21% of GDP to 15%). Since that date, subnational government expenditure is quite stable, the transfer of spending responsibilities in the social sector in 2007 and 2015 being counterbalanced by public finance consolidation measures and management reforms (OECD, 2014[8]).

Figure 3.1. Decentralising or recentralising trends in the OECD over 1995-2016
Changes in subnational government expenditure as a percentage of GDP and subnational government between 1995 and 2016

Note: Iceland 1998-2016; Japan: 2005-16, Mexico: 2003-16. No data for Australia, Chile and Turkey due to lack of time-series. OECD30 average is unweighted and does not include Australia, Chile, Japan, Iceland, Mexico and Turkey. Information on data for Israel:

Source: OECD calculations based on OECD national accounts (accessed on 9 October 2018).

An increase in revenue decentralisation

At the OECD level, the share of grants in total revenue was relatively stable on average between 1995 and 2016. There have, however, been two main stages during this period, before and after the crisis (OECD/KIPF, 2016[9]). In the 1990s until about 2008, a widespread policy of decentralisation of responsibilities in areas such as education, social protection, transport infrastructure, utilities, etc., was funded mainly by central government transfers, while subnational tax revenues remained largely stable. This resulted in an increase in vertical fiscal imbalance (the difference between subnational government own revenues and their spending obligations). An ever-growing grant system became the mechanism used to respond to higher subnational spending. Increased use of transfers was also employed to prevent regional disparities, correct horizontal imbalances across jurisdictions and meet central government requirements in service delivery (Bergvall et al., 2006[10]).

At the country level, from 1995 to 2005, the share of transfers grew mainly as a response to the widespread policy of decentralising expenditures. During the economic and fiscal crisis, grants increased considerably in some countries as a means to support recovery plans, but this did not last. After the crisis, public finance consolidation plans in many countries resulted in freezes and cuts in central government transfers. The highest decrease of grants as a share of subnational revenues between 2005 and 2016 was observed in France, Greece, Ireland, Portugal and Slovenia. As a result, despite the up and down swings of the last 21 years, today the composition of subnational government revenue is close to where it stood in 1995, on average in the OECD. At the country level, over the entire 1995-2016 period, the share of grants in total subnational revenues increased the most in the Denmark, Latvia, New Zealand and the Slovak Republic. The highest decreases were registered in the Czech Republic, Hungary, Ireland and Italy.

Figure 3.2. Grants as a share of subnational revenues in 1995, 2000 and 2016

Note: Ireland: 1998-2016; Japan: 2005-16; Mexico: 2003-16. No data for Australia, Chile, Turkey. OECD average is unweighted.

UWA: unweighted average for all countries

Source: OECD own elaboration based on OECD national accounts (data extracted in October 2018).

On the tax revenue side, tax revenues have increased slightly both in GDP and in total public tax revenues. Tax revenues encompass shared and own-source taxes, which does not allow for assessing progress in tax autonomy. Between 1995 and 2016, subnational tax revenues increased – or decreased – significantly in several countries as a percentage of total public tax revenue and GDP. These changes typically reflect economic performance since taxes such as personal income (PIT), corporate income (CIT), VAT, property transaction, consumption, construction activity, etc., are sensitive to economic fluctuations. The changes also reflect tax reforms indirectly affecting subnational governments (e.g. changes in national shared PIT or CIT) or directly related to subnational governments.

Several tax reforms took place over the past 21 years that aimed at increasing the importance of taxes in subnational funding, either by allocating larger shares of national taxes to subnational governments and/or by giving more taxing powers to subnational governments (ability to create local taxes, to determine rates and bases and to grant tax allowances or reliefs). These tax reforms resulted in significant increases in subnational tax revenues in Belgium (2002 and 2014), the Czech Republic (2005), Poland (2004), Slovenia (2009) and Spain (in 2002, 2009 and 2012) and. In Italy, after the 1998 Bassanini Reform, tax revenue increased vastly from 25% in 1997 to 41% of subnational revenue after the 2009 reform.

Figure 3.3. Changes in grants and tax revenue as a share of subnational government revenue
Changes in percentage points between 1995 and 2016

Note: Ireland: 1998-2016; Japan: 2005-16; Mexico: 2003-16. No data for Australia, Chile, Turkey. OECD average is unweighted.

Source: OECD own elaboration based on OECD national accounts (data extracted in October 2018).

In contrast, the share of subnational tax revenue contracted in Denmark (especially since the 2007 local government reform and the abolition of counties, which previously benefited from tax revenues), Hungary (fiscal recentralisation reform, Latvia, Norway (abolition of the national corporate income tax as a local tax in 1999), the Slovak Republic (2005 reform) and Switzerland.

However, these figures do not offer a fair view of the changes as tax revenue figures distinguish between shared taxes and own-source taxes, for which subnational governments have a certain leeway over rates and bases. In France for example, the 2010 local finance reform did not affect the share of tax revenue in local revenue or GDP, but significantly diminished the share of own-source taxes, leading to less tax autonomy. In other countries, the reverse is also true: stability in terms of tax revenue may have been observed over the years although there was an increase in tax autonomy resulting from the introduction of new own-source local taxes and the broadening of local decision-making power for setting rates or based on existing local taxes (e.g. property tax).

To provide a complete picture, it is worth looking at the tax autonomy indicators developed by the OECD Fiscal Network. They show that from 1995 to 2011, tax autonomy increased, at the expense of tax-sharing systems (OECD/KIPF, 2016[9]). In Denmark and France, fiscal reforms led to a decrease in subnational revenue autonomy, while in Finland, Italy, Japan, Korea, Portugal and Turkey local revenue autonomy increased (OECD, 2017[11]).

Figure 3.4. Decentralising or recentralising trends in the OECD over 1995-2016
Changes in subnational government tax revenue as a percentage of GDP and as a percentage of public tax revenue

Note: Iceland 1998-2016; Japan: 2005-16; Mexico: 2003-16. No data for Australia, Chile and Turkey due to lack of time-series. OECD30 average is unweighted and does not include Australia, Chile, Japan, Mexico and Turkey. Information on data for Israel:

Source: OECD calculations based on OECD national accounts (data extracted in October 2018).

A modest increase in subnational spending power from 1995 to 2014

Financial expenditure shares often do not accurately reflect subnational spending power, as subnational spending is generally highly regulated or otherwise influenced by the central government. Using results from the recent spending power questionnaire sent to member countries of the OECD Fiscal Federalism Network, the OECD constructed institutional indicators that compare subnational spending power across countries and policy sectors (OECD, forthcoming[12]).

According to the results, subnational spending authority is relatively low, on average. Spending power is quite balanced across sectors but is lower in health and long-term care, and higher in social housing, transport and primary and secondary education (referred to as education).

While the data collected from single questionnaires do not enable comparisons over time, the OECD Fiscal Federalism Database includes useful information for descriptive purposes. For instance, the evolution of taxing power from 1995 to 2014 can be described using data on subnational tax revenue as a percentage of GDP and subnational tax revenue as a percentage of total tax revenue.

The average share of subnational tax revenue increased slightly from 1995 to 2014 (Figure 3.5). The pattern of taxing power changes was more complex, including ups and downs depending on years and types of subnational government in question (the database has separate figures for states and local governments). On average, the tax revenue share of subnational governments in all tax revenue rose by 0.9 percentage points, by 1.1 percentage points for the state level and by 0.43 percentage points for local governments.

Figure 3.5. A slight increase in SNG tax autonomy indicators during 1995-2014
Difference of SNG tax revenue shares between years 1995 and 2014, percentage points

Note: Author’s calculations. Information on data for Israel: Local governments in the United States have a wide variety of taxing powers but it is not possible to identify the share of each. The unweighted average for 1995 and 2014 applies to the subnational revenue shares in the 35 OECD countries. The unweighted average for States applies only to the ten countries reporting state or regional data. Italy and Spain are considered as regional countries for the purpose of the tax autonomy indicators.

Source: OECD (n.d.[13])OECD Fiscal Decentralisation Database

Enhanced local autonomy as measured by the LAI

As highlighted in Chapter 2, financial data are not sufficient to assess local autonomy. There are other elements, such as the extent to which municipalities are able to decide on the kind or type of services they wish to provide, how they want to organise themselves, and whether they have an influence on decisions taken at a higher government level. The following section looks at the trends in the different dimensions of local autonomy in 39 European countries (Keuffer and Ladner, 2018[14]).

The Local Autonomy Index consists of seven dimensions:

  1. 1. Legal autonomy measures the extent to which the existence of municipalities is constitutionally guaranteed and whether or not municipalities can, for example, be amalgamated against their will.

  2. 2. Policy scope describes the range of services for which municipalities are responsible.

  3. 3. Political discretion additionally asks whether municipalities also have some decisional power while fulfilling these tasks.

  4. 4. Financial autonomy means that municipalities have their own financial resources, can collect tax and decide on their base and their rate and are able to borrow money.

  5. 5. Organisational autonomy describes the possibility to organise and staff their administration and to decide on features of their political system.

  6. 6. Non-interference is related to the vertical relations with the higher levels of state and consists in the way supervision is organised and whether financial transfers are unconditionally granted.

  7. 7. Access captures whether municipalities can influence higher-level decisions.

The LAI concerning the legal dimension suggests that lawmakers in Europe gradually seem to be strengthening the level of autonomy granted to local authorities (Table 3.1). With respect to policy scope, there is still substantial variation across European countries, with the Nordic countries and some Central European states at the top end of the scale, and some Black Sea countries as well as Ireland and the United Kingdom at the other end. Policy scope also varies across the policy fields considered (i.e. education, social assistance, health, land-use, public transport, housing, police and caring functions). Policy responsibility is most extensive in the fields of land-use (zoning, building permits), school buildings, housing, caring functions and public transport. In many European countries, it is municipalities are that are mostly in charge of these functions. It is more unusual for municipalities to have responsibility for the police, what actually goes on inside school buildings (e.g. educational programmes, the hiring and paying of teachers) or in health services.

Financial autonomy witnesses a huge variation, both across time and among countries. For most of the 1990-2014 period, the dominant trend was an increase in financial autonomy. This was especially true for post-communist countries in Central and Eastern Europe, which started out with a very low level of financial autonomy. There was also an increase in some countries from other parts of the continent (e.g. Italy and Malta). With the 2008 financial crisis, the overall trend was partially reversed, most visibly in changes related to borrowing regulations and in countries that were most severely hit by the crisis. It is interesting to note that the LAI shows no evidence that a large scope of functions allocated to local governments generates demands to tighten control over municipal finance. Rather the opposite seems to be the case: taxation and borrowing autonomy are often higher in countries with more functional decentralisation.

According to the LAI, most European local government enjoy quite high levels of organisational autonomy. Municipalities are usually able to elect their local executive directly and have some leeway when it comes to organising their local administration. In some countries, municipalities can decide on elements of their electoral system, such as the number and size of their electoral districts, whether they prefer majority elections or proportional representation, and the form or the size of their local executive (e.g. in the Czech Republic, Denmark, Estonia, Iceland, Norway, Sweden, Switzerland). In most countries, however, these parameters are set by national legislation. As for local administration, in most countries, local government has the freedom to hire its own staff, fix the salaries of employees, choose their organisational structure and establish legal entities and municipal enterprises. There are, however, also countries where the local administration is more directly organised and administered by the central state (e.g. France). If there have been changes in the degree of organisational autonomy, most of them took place in the 1990s. For many countries, particularly for those in Eastern and Central Europe, active democratic reforms and a transformation of the prevailing political culture were consolidated in this period.

The results of the LAI point out a general increase in autonomy between 1990 and 2009. A similar pattern is shown by the OECD Fiscal Federalism Database indicators. This increase has taken place especially in the new democracies of Central and Eastern Europe, such as Albania, Bulgaria and Romania. However, there are also countries, such as Hungary, that have pursued centralisation policies over the last few years. The LAI shows that in 2014, Switzerland and the Nordic countries (Denmark, Finland, Iceland, Norway and Sweden) rank among the countries with the highest degree of autonomy, together with Germany, Liechtenstein and Poland (Table 3.1 and Box 3.1).

Table 3.1. Local Autonomy Index (LAI), country ranking 2014, 1990 and changes





Countries (next)








































































































United Kingdom












Czech Republic








































Slovak Republic








Note: Countries sorted by 2014. For Albania, Latvia, Malta, Romania and Ukraine, there is no data for 1990; the first years of measurement for these countries are: 1992, 1991, 1993, 1992 and 1991 respectively.

Note by Turkey: The information in this document with reference to “Cyprus” relates to the southern part of the Island. There is no single authority representing both Turkish and Greek Cypriot people on the Island. Turkey recognises the Turkish Republic of Northern Cyprus (TRNC). Until a lasting and equitable solution is found within the context of the United Nations, Turkey shall preserve its position concerning the “Cyprus issue”.

Note by all the European Union Member States of the OECD and the European Union: The Republic of Cyprus is recognised by all members of the United Nations with the exception of Turkey. The information in this document relates to the area under the effective control of the Government of the Republic of Cyprus.

Source: (Keuffer and Ladner, 2018[14]).

Box 3.1. Country-specific patterns of local autonomy

More revealing than a simple overall ranking are country-specific patterns of local autonomy. Finland, for example, has a “full” profile, reaching high levels of autonomy in all seven dimensions, whereas the opposite is the case for Ireland. By contrast, high legal autonomy with less autonomy in the other dimensions is characteristic among the newer Eastern democracies such as Bulgaria, the Czech Republic, Estonia and Romania. The latter two distinguish themselves from the other countries by their high degree of organisational autonomy.

France, after its increased decentralisation efforts in the 1980s, demonstrates a substantial degree of autonomy in almost all dimensions, with the exception of features of the local political system and administrative organisation. In Switzerland, municipalities are autonomous in financial and organisational affairs and enjoy legal protection, but they are more limited in their ability to decide on their policies. This is related to their small size and the Swiss form of federalism, which allocates political discretion to the cantons. The much larger German municipalities – despite the federalist structure of the country – are more autonomous with respect to policy scope and political discretion.

Hungary, Slovenia and Ukraine share a similar pattern: a lack of financial autonomy and deficits with respect to access and non-interference. They score average on autonomy on the other four dimensions. Interesting to note are also the similarities between Spain and the United Kingdom. In the former, however, organisational autonomy is low and financial autonomy is high whereas in latter the opposite is the case. The patterns for Ireland finally show very low levels of autonomy in virtually all dimensions.

Source: Keuffer, N. and A. Ladner (2018[14]), The Local Autonomy Index Project – Extent, Patterns and Effects of Local Autonomy in Europe.

Preliminary analysis suggests that the LAI is positively related to trust in local and regional government, even when other variables are controlled (Keuffer and Ladner, 2018[14]). On the output side, the analyses show two significant correlations when observing the average size of the municipalities and post-communist countries: between local autonomy and a low degree of corruption plus a high degree of happiness. The LAI analyses also show that financial autonomy is the dimension that almost consistently correlates most strongly with the different performance measures. Quite often, there are also positive links with the policy scope of local government and political discretion on the one hand, and performance indicators on the other hand. These results indicate a need for additional analysis. Developing these indicators further is an essential step to facilitate high-quality empirical analysis of the effects and effectiveness of decentralisation.

Upscaling governance: The rising role of regions and metropolitan areas

If decentralisation implies a strengthening of local autonomy, it also goes hand in hand with an upscale in subnational governance through municipal co-operation, metropolitan governance and the strengthening of regions. This section provides evidence behind these trends at the inter-municipal, metropolitan and regional levels, and highlights some policy implications.

Municipal consolidation and co-operation

Municipal fragmentation, which can result in an inefficient provision of local services and raises issues of equity in access to services generally and of equivalent quality, has been the driver of policies that encourage or impose amalgamation. It has also engendered policies that foster inter-municipal co-operation (OECD, 2017[11]) as a way to generate economies of scale, efficiency gains and cost savings.

Municipal mergers have taken place in most OECD countries over the last decades. In some countries, there have been successive waves of reform. In others, it has been a gradual process over a long period, which may still be underway. The global crisis acted as a catalyst to reactivate or introduce municipal amalgamation policies. Recent reforms took place in Greece (2011), Turkey (2012-2014), Ireland (2014) and Estonia (2017) where the number of municipalities decreased from 213 to 79 (14 of which are urban and 65 rural) following the administrative reform completed in October 2017 (OECD, 2018[15]). Municipal mergers are still ongoing as a piecemeal process in several countries such as in Finland, France, Italy, Luxembourg, the Netherlands and Norway. Some federal countries also introduced mergers under the leadership of federated states e.g. Thurgovie, Fribourg, Vaud, Tessin, Grisons in Switzerland, New South Wales and South Australia in Australia, Styria in Austria, Manitoba or New Brunswick in Canada, and Brandenburg and Saxony-Anhalt in Germany.

Figure 3.6. Municipal mergers in selected OECD countries from 1950 to 2017

Source: OECD (2017[11]), Multi-level Governance Reforms: Overview of OECD Country Experiences,, updated.

Reasons for mergers are numerous, including the need to adapt to demographic change (e.g. ageing, migration, shrinking or expanding populations) and to reduce the mismatch between obsolete municipal administrative boundaries and socio-economic functional areas. Reasons are also linked to the need to implement economies of scale and scope, generate cost savings and internalise spillovers in the provision of local public services, to increase municipal administrative capacities, to improve the quality and quantity of municipal infrastructure and services and more generally, to improve the governance, professionalism and efficiency of municipal management (OECD, 2017[11]). However, despite incentives to merge, the implementation of merger policies has often faced resistance, leading some countries either to impose municipal amalgamations in a compulsory way, to temper initial ambitions or to abandon the project altogether.

Today, inter-municipal co-operation (IMC) is widespread in the OECD, and firmly rooted in European and OECD municipal management practices. It is relevant in all countries regardless of municipal size, offering benefits to rural and metropolitan areas alike (Chapter 4). There have been different drivers for this progress in inter-municipal co-operation, especially since the global crisis (Figure 3.7).

Figure 3.7. Drivers for inter-municipal co-operation in the OECD

Legal frameworks and policies supporting inter-municipal co-operation have been significantly enhanced over the last 15 years (Table 3.2). Inter-municipal arrangements are now extremely diverse and reflect varying degrees of co-operation. There are different formats for inter-municipal co-operation in the OECD, ranging from the softest – single or multi-purpose co-operative agreements/contracts (e.g. shared services arrangements or shared programmes in Australia, England [United Kingdom], Ireland, New Zealand) – to the strongest forms of integration (e.g. supra-municipal authorities with delegated functions in France, Portugal and Spain). Between the two, co-operation covers a range of forms and areas, from co-operation focused on technical issues to strategic co-operation for economic and social development (OECD, 2017[11]).

OECD countries generally have chosen to start first with a private law model. This is usually based on the freedom of local authorities to opt for certain formulas, such as contracts, associations and commercial enterprises. The next step is often a move to a public law model. The public model means that co-operation is regulated in some detail by public laws, including the contractual and financing arrangements, the type of delegated functions, the governance structure, the supervision and control, etc.

Table 3.2. Progress of inter-municipal co-operation in the OECD


Legal frameworks and policies


Regional Organisations of Councils (ROCs), a major type of multi-purpose co-operation arrangement among councils for over 70 years, have been substantially restructured, reformed or have disappeared entirely in some states. Other types of IMC exist: Regional Local governments (Western Australia); Regional Subsidiaries (South Australia); County Councils (NSW), etc. Shared services arrangements promoted in both state and local government throughout Australia.


IMC designed and implemented by individual states but the Constitution states that municipalities may join forces by agreement or by law to form inter-municipal organisations (Gemeindeverbände). A 2011 federal constitutional amendment further reinforced the ability of municipalities to engage in inter-municipal associations.


Flemish and Walloon Regions are currently encouraging “supra-municipalities”. In Flanders, the existence of five legal forms of IMC: “interlocal” association, project association, service providing association, association with a clear assignment, association with a clear assignment and private sector participation.


Special agencies, joint boards and commissions are popular to provide specific services to groups of municipalities. Some provinces can have specific policies promotion IMC such as New Brunswick.


A 2009 law allowed municipal associations to obtain legal personality under private law. In 2011, a new law provided the impetus to create municipal associations.

Czech Republic

IMC promoted by the 2000 Act on Municipalities under the form of voluntary municipal association and micro-regions (around 790 IMC structures active in the field of education, social care, health, culture, environment, tourism).


Voluntary IMC for example in “Business Regions”.


Existing framework for voluntary IMC to perform non-mandatory tasks on behalf of local authorities defined in the 1993 Local Government Organisation Act (non-governmental organisations, municipal enterprises, etc.). Enhanced IMC promoted by the recent territorial reform.


Special purpose joint authorities (184 currently) producing joint services several municipalities, mostly in healthcare and education will be touched by the ongoing local government reform, creating regions and transferring some municipal functions to them.


Different categories of IMC structure with own-source tax (EPCI à fiscalité propre), including since the changes introduced by the 2014 MAPTAM law: “communities of municipalities” in rural areas, “town communities”, 11 “urban communities” for urban areas and metropolis (more than 400 000 inhabitants). Recent 2015 Law NOTRE set up a minimum threshold for inter-municipal groupings.


IMC is strongly encouraged by Länder, in particular for some standard local services such as waste management, sewage, water or transport (Zweckverbände and municipal associations).


IMC structures with single or multiple tasks authorised by the municipal code but many of them are inactive.


IMC is encouraged through the 1997 Act on the Associations and Co-operation of Local Government. The 2011 Cardinal Law on Local Governments stated compulsory pooling of administrative services for municipalities under 2 000 inhabitants in joint offices or districts or micro-regions.


IMC takes place on a regional basis through regional boards, regional federations and economic development agencies co-owned by the municipalities (Local Government Act n°45/1998). IMC mandatory for small municipalities (less than 8 000 inhabitants) following the decentralisation of social services for disabled people.


Shared services programme for waste management regulation and shared services currently being developed.


Existence of municipal associations


IMC promoted since 1990 in particular through municipal unions, mountain communities. 2014 Law no. 56 encouraged IMC by strengthening the municipal unions, establishing a minimum threshold is established and extending their scope of tasks. IMC compulsory for small municipalities (less than 5 000 inhabitants). This law also transformed the provinces into IMC bodies, including the 14 metropolitan cities.


IMC takes place through voluntary Partnership Agreement established under the Local Autonomy Act to be implemented in 2014.


Local government may establish intergovernmental corporate authorities but little use.


Around 75 Intercommunal associations with quite a wide range of activities.


Since 1999, municipalities can create an inter-municipal association and formalise inter-municipal agreements for the joint supply of public services such as water and sewage, public security and public transport.


Around 700 IMC structures in 2010 concluded under the Joint Regulations Act (Wet Gemeenschappelijke Regelingen or WGR Act) and since 2004 with the Work and Social Security Act (creation of shared services centres e.g. for social services) and functional regions for safety and healthcare.

New Zealand

IMC and shared services between local authorities encouraged by the 2013 Local Government Act 2002 Amendment Bill.


Since the 1992 Local Government Act and Intermunicipal Companies Act of 29 January 1999, IMC agreements and inter-municipal bodies are widespread. In 2006, an amendment to the Local Government Act widened the range of delegated tasks to IMC bodies to a “host municipality” through an agreement.


Introduction of the “territorial contracts” in 2014 which are intended to strengthen partnership and improve co-ordination of territorially oriented activities of various stakeholders.


IMC encouraged since 2003 laws creating intermunicipal communities (comunidades intermunicipais). In 2013, a new law created 23 compulsory intermunicipal communities regrouping previous urban communities, intermunicipal communities for general purpose and some metropolitan areas created in 2003 and abolished in 2008. Development of multi-level contracts: the Partnership Agreements with the European Commission are used to promote IMC.

Slovak Republic

Municipalities co-operate in the framework of voluntary “joint municipal office” which are multi-purpose and implementing co-ordination arrangements covering 21 different domains. The 2012-203 ESO Programme reinforced the role of joint municipal office for the purposes of managing functions delegated from the state.


2005 amendments to the Financing of Municipalities Act provided financial incentives for voluntary joint municipal administration.


Mancomunidades are voluntarily established entities to carry out joint projects and provide common services. Law 27/2013 promotes the integration or co-ordination of municipal services through financial incentives.


IMC rules defined by the Local Government Act establishing contracts, “common committee” to run joint services in healthcare or education and “municipal federation” (kommunalförbund).


IMC framework and policies are designed and implemented by cantonal constitutions and laws


Municipalities can provide jointly some services, especially in rural areas, through municipality unions and unions for irrigation.

United Kingdom

In England, development of “shared Service Agreements” involving collaboration between two or more local authorities and promotion of combined authorities in urban areas. In Scotland, councils are able to set up joint board or joint committee organisations with other councils to provide a service across a combined area of the participating local authorities.

United States

IMC designed and implemented by individual states; 51 146 formal entities in 2012 (schools districts, transport districts, water management, fire services) undergoing a consolidation process. Formal shared services arrangements commonly directed by state government have a 20- to 30-year history of development.

In terms of financing, inter-municipal bodies’ structures are most often financed through contributions from municipality members but they usually complement these subsidies with other revenue sources related to the services they provide (user charges and fees). They can also receive grants from the central government. This is a way for the central government to favour co-operation, and has been the practice in several OECD countries (e.g. Estonia, France, Norway and Spain). France is the only OECD country that has systematically set up public institutions for inter-communal co-operation with taxing powers, i.e. able to raise their own sources of tax revenue (EPCI à fiscalité propre).

Figure 3.8. From soft agreements to more formalised forms of co-operation in the OECD

Source: Adapted and completed by the OECD based on

Metropolitan governance

Metropolitan governance reforms aim at addressing the issue of municipal fragmentation in large urban areas. Efficient metropolitan governance has become a hot topic in many countries. Administrative borders in metropolitan areas, based on historical settlement patterns, no longer reflect current human activities or economic and social functional relations (OECD, 2015[16]). Enhancing co-operation and the co-ordination of public policies on a metropolitan-wide basis, in particular with regard to the provision of public infrastructure and services, aims to improve the quality of life and international competitiveness in large cities.

The number of metropolitan governance authorities has increased considerably and there has been renewed momentum in the number of metropolitan governance bodies created or reformed since the 1990s, against the backdrop of the early 1990s recession and the 2008 financial crisis. Currently, around two-thirds of the metropolitan areas in the OECD have a metropolitan governance body (Ahrend, Gamper and Schumann, 2014[17]). The additional responsibilities given to metropolitan areas are often linked with infrastructure and planning tasks, such as public transport, environment, and spatial planning, as well as services targeted at local business.

Recent years have seen more and more OECD countries adopting differentiated municipal governance for metropolitan areas. Capital districts with special autonomy often started out in the mid-20th century with lower levels of authority because of restrictions on representation or other elements of self-rule.

Figure 3.9. Number of metropolitan governance structures created or reformed in the OECD, by decade

Source: OECD update based on Ahrend, R., C. Gamper and A. Schumann (2014[17]), “The OECD Metropolitan Governance Survey: A Quantitative Description of Governance Structures in large Urban Agglomerations”,

The latest update of the Regional Authority Index (RAI) (2010-16) included a focus on metropolitan and urban governance (see Box 3.2). Metropolitan and urban government in 45 countries that fulfil the criteria for regional government all contribute to differentiation in subnational governance, but they do so in different ways. Although metropolitan and urban government constitute an additional tier of subnational government in only part of the country, they often exercise authority over a large part of the population. For example, in France, the communautés urbaine (urban communities) and métropoles (metropolises) involve 1 485 out of a total of more than 36 000 (4.1%) municipalities and exercise authority over about 27.7 million inhabitants which constitutes 43% of the total population. Italian metropolitan cities (città metropolitana) are home to almost 22 million out of a total of 59 million people (37% of the total population). Seven combined authorities in the United Kingdom involve 44 constituent local authorities and about 13.5 million inhabitants, which is around 21% of the total population. In some countries, metropolitan and urban government affect almost the whole population. In Portugal Continental (mainland), inter-municipal communities and the metropolitan areas of Lisbon and Porto (Área Metropolitana de Lisboa/do Porto) involve all 278 municipalities (out of a total of 308 municipalities including Autonomous Regions of Azores and Madeira), 90% of the total municipalities, and comprise almost 9.8 million out of a total of 10.3 million citizens (95%)

A second way that metropolitan governments may contribute to differentiation in subnational governance is when they have additional competencies when compared to their “peers” within the same government tier. For example, metropolitan cities in Italy (città metropolitana) take over the competencies of provinces and are given additional responsibilities for local police services, roads, transport, and spatial and urban planning. The metropolitan city mayor is directly elected. In comparison to provinces, metropolitan cities score higher score on the RAI- policy scope and executive indicators but they have similar scores for institutional depth, fiscal autonomy, and borrowing autonomy. In Hungary, cities with county rights are allowed to borrow with prior approval by the central government whereas counties have no borrowing autonomy.

Box 3.2. Metropolitan governance as measured by the Regional Authority Index

The Regional Authority Index (RAI) defines metropolitan and urban government in the same way as a regional government, which means that:

  1. 1. It is an intermediate tier of government in between local, municipal tier and national government. A sub-metropolitan or sub-urban tier can consist of councils or assemblies established in city districts or in the member municipalities.

  2. 2. It is a multi-purpose and not single- or specific-purpose government.

  3. 3. It legally exercises competencies and is not a collaboration purely based on a voluntary basis. This means that the competencies of metropolitan and city government are laid down in legislation, either in a specific law or a chapter in a local/regional government law.

  4. 4. It has an average population size of 150 000 people or more across units.

Source: Schakel, A. et al. (2018[18]), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

A third way in which metropolitan and urbanised governments contribute to differentiation in subnational governance is through special autonomy arrangements that only apply to the capital. Seven capital cities with their own law and which fulfil the criteria of regional government have been identified: Barcelona, Brussels, Bucharest, Copenhagen, Helsinki, Paris and Prague. What sets these capitals apart from other capitals with special laws is that they are a general-purpose government that exercises significant additional authority when compared to other regional government units. For example, Prague is both a municipality and a region and, in contrast to other regions which are completely reliant on intergovernmental grants, Prague can set the property tax rate in its capacity as a municipality. Similarly, Bucharest is a municipality with county rights that, before 1999, had more fiscal and borrowing autonomy when compared to other counties thanks to its legal status as a municipality.

There are 26 metropolitan and urban governments, showing that this type of subnational governance is a relatively recent phenomenon. The general trend in regional authority shows a significant increase from the 1970s onwards. Most subnational and urban governments were established in the 1990s and 2000s. Out of the 11 metropolitan and urban governments that were created before 1990 all but 2 concern one particular area or territory. Out of the 16 metropolitan and urban governments that were set up in or after 1990, 8 concern a singular entity and 8 involve multiple government units.

All metropolitan and urban governments have responsibilities relating to industrial promotion, environmental planning, garbage disposal, public transport, regional spatial planning, regional economic development, recreation, regional parks, tourist promotion, traffic planning, traffic regulation and water supply. Nine out of 24 metropolitan and urban governments score 2 on policy scope because they have (limited) competencies in cultural-education policy (e.g. cultural, sport and recreational facilities, inter-municipal libraries, museums and school buildings) and/or in welfare policy (i.e. hospitals, public health, social housing or social services).

Another notable feature of metropolitan and urban government is their limited fiscal and borrowing autonomy. Nine metropolitan and urban governments rely completely on municipal transfers and user fees whereas 15 can set the rate of a property tax.1 In addition, 6 metropolitan and urban governments cannot borrow and 15 can only borrow under strict rules and with ex ante approval from the central (or higher regional) government. In general, the fiscal capacity and budgetary autonomy of metropolitan and urban government is particularly limited, which makes them more reliant on intergovernmental transfers from the participating municipalities and cities and from higher regional and national governments. This may not only be detrimental for their autonomy but also complicates budgetary negotiations.

Rising role of regions

The Regional Authority Index (RAI) presents a useful way to explore trends in decentralisation across a large number of countries. This index is a comprehensive attempt to measure the real degree of power of intermediate governments, going beyond fiscal indicators. The RAI specifically focuses on regional government, which is defined as an intermediate tier of government between the lowest, local tier and national government, with at least 150 000 inhabitants per regional unit on average. This indicator traces regional authority across 10 dimensions in 81 countries between 1950 and 2010 (Hooghe et al., 2016[2]; Hooghe, Marks and Schakel, 2010[19]). The RAI’s ten dimensions include fiscal autonomy and borrowing autonomy, but also lawmaking and executive control (Box 3.3; Annex A).

The RAI consists of two components, self-rule and shared rule, each of which has five dimensions. Table 3.4 displays the number of reforms for self-rule and shared rule for each of the 10 dimensions in 81 countries between 1950 and 2010. Two observations are particularly important: i) the total number of reforms affecting self-rule is more than three times larger than the total number of reforms affecting shared rule; ii) a closer look reveals that fiscal indicators have been least subject to reform. Reforms are three times less likely to involve fiscal autonomy and control, and borrowing autonomy and control, rather than other dimensions of authority, policy scope or representation (Figure 3.10).

Box 3.3. Self-rule and shared rule in the RAI

Self-rule is the authority exercised by a regional government over those who live in the region. Self-rule taps into the extent to which a regional government is autonomous rather than deconcentrated (institutional depth), the range of policies for which a regional government is responsible (policy scope), the extent to which a regional government can independently tax its population (tax autonomy), the extent to which a regional government can borrow (borrowing autonomy), and the extent to which a region has an independent legislature and executive (representation).

Shared rule is the authority exercised by a regional government or its representatives in the country as a whole. Shared rule can be exercised through regional representatives in an upper chamber of the national parliament and through intergovernmental meetings with national and other regional governments. Similar to self-rule, regions can have shared rule across a range of dimensions:

  • the extent to which regional representatives co-determine national legislation (law making)

  • the extent to which regional executives co-determine national legislation (executive control)

  • the extent to which regional representatives co-determine the distribution of national tax revenue (fiscal control)

  • the extent to which a regional government co-determines subnational and national borrowing constraints (borrowing control)

  • and the extent to which regional representation co-determines constitutional change.

Annex A provides further detail on the scores for each of the ten dimensions.

Source: Hooghe, L. et al. (2016[2]), Measuring Regional Authority: A Postfunctionalist Theory of Governance, Oxford University Press, Oxford.

Table 3.3. Number of reforms broken down into ten dimensions of regional authority







Institutional depth



Law making



Policy scope



Executive control



Fiscal autonomy



Fiscal control



Borrowing autonomy



Borrowing control






Constitutional reform






Shared rule



Note: Shown are the number of positive and negative reforms for 5 dimensions of self-rule and 5 dimensions of shared rule for 81 countries for 1950-2010. A reform is included when there is at least a 0.1 change in the magnitude of a country RAI-score in a particular year.

Source: Schakel, A. et al. (2018[18]), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

Figure 3.10. Type of reforms strengthening self-rule in regions

Source: Schakel, A. et al. (2018[18]), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

All regions are concerned about the increase in the RAI (Box 3.3). Figure 3.11 displays average RAI-scores for American, Asian and European countries since 1950. A striking observation is a clear trend of increasing regional authority across the three groups of countries and this trend is especially noticeable from 1970 onwards. Average regional authority was 55% higher in 2010 than in 1950. Of the 81 countries covered by the RAI, 52 experienced a net increase in the degree of regional authority and only nine experienced a net decline.2

Box 3.4. RAI score in Latin America, Asia and Europe

The average RAI-score in 1970 was 10.3 for America, 15.5 for Asia, and 17.3 for Europe. In 2010 the average RAI-score increased to 22.6 in America, to 29.3 in Asia, and to 27.1 in Europe. An increase of 10 to 14 points indicates that, on average, each country included in the RAI has introduced a fully equipped intermediate tier of government. A 12-point increase in the RAI is equal to establishing a general purpose regional tier subject to central government veto (+2 points), which has authoritative competency in economic and cultural-educational policy (+2 points), which can set the base and rate of a minor tax (+2 points), which can borrow without prior authorisation by central government (+2 points), and which has a directly elected assembly (+2 points) and a regional executive appointed by the regional assembly or directly elected (+2 points).

Source: Schakel, A. et al. (2018[18]), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

The motivation for regionalisation reforms varies across countries. The size of the country matters: large countries tend to have more layers of subnational governments (OECD-UCLG, 2016[20]). However, many countries of a relatively modest size have also introduced or strengthened a regional level in recent decades. The main objectives behind regionalisation reforms are an upscale in governance to generate economies of scale for public service provision, for example in the health or public transport sectors. Regions may also better respond to widening functional labour markets. Other objectives are to improve co-ordination between municipalities and intermediate levels of government in such areas as infrastructure delivery, spatial planning and land-use, for example. Larger regions are also expected to be more competitive: they have higher critical mass, more resources to implement effective regional development strategies, and the ability to foster intra-regional co-ordination and to implement more integrated territorial planning. They may better target regional comparative advantages through access to local knowledge, compared to the national government or to fragmented local governments.

Several countries have created new regions, notably in East European countries in the context of EU enlargement. Others have strengthened existing regions: this is the case in recent or current reforms in Nordic countries, France or Italy. In several Nordic and Central and East European countries, responsibilities such as higher education, specialised healthcare and regional public transport were reassigned from the municipal and the central government levels to a newly created regional level. However, in recent years some European countries have also gone in the opposite direction and have reduced the role of regions. This is the case in Denmark with a reduction in the responsibilities assigned to regions, for example, or in Hungary through recentralisation reforms.

Figure 3.11. Regionalisation in America, Asia and Europe since 1950

Note: Shown are average Regional Authority Index scores for 29 American, 11 Asian and 41 European countries. America: Argentina, Bahamas, Barbados, Belize, Bolivia, Brazil, Canada, Chile, Colombia, Costa Rica, Cuba, the Dominican Republic, Ecuador, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, El Salvador, Suriname, Trinidad and Tobago, Uruguay, the United States, and Venezuela. Asia: Australia, Brunei, Indonesia, Japan, Malaysia, New Zealand, the Philippines, Singapore, South Korea, Thailand, and Timor Leste. Europe: Albania, Austria, Belgium, Bulgaria, Bosnia and Herzegovina, Croatia, Cyprus,1 the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Kosovo,2 Latvia, Lithuania, Luxembourg, Malta, Montenegro, the Netherlands, Norway, Poland, Portugal, Romania, Russia, Serbia and Montenegro (until 2006), Serbia, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, the United Kingdom and Turkey.

1. See Table 3.1 for notes regarding Cyprus.

2. This designation is without prejudice to positions on status, and is in line with United Nations Security Council Resolution 1244/99 and the Advisory Opinion of the International Court of Justice on Kosovo’s declaration of independence.

Source: Schakel, A. et al. (2018[18]), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

Table 3.4. Main objectives of regionalisation reforms

Broad policy objective

Governance and management objectives

Solve current challenges


Make economies of scale in public service provision

● Improved efficiency of health system provision

● Economies of scale in labour market services, public transport, infrastructure

● Internalisation of spillovers, better quality public service provision

● Improved co-ordination between local governments

● Fragmentation of responsibilities for public service delivery in many policy areas (infrastructure, transport, healthcare, housing, etc.)

● Risk of creating clumsy constructions

● Optimal size debate: very hard to identify an optimal size for efficient public service delivery

● Efficiency-related counterarguments, presented by those who believe that smaller units may be more efficient and less bureaucratic since they have better local knowledge, can choose measures more adapted to the needs of their clients and are better able to adjust service provision to variations in local demand

Enhance competitiveness and regional growth

● Bigger regions would have higher critical mass, more resources to implement effective regional development strategies, the ability to foster intra-regional co-ordination and to implement more integrated territorial planning

● Better access to local knowledge, remedy to asymmetries of information

● For public goods with strong local/regional externalities – related to regional development such as innovation, labour market policy as spatial planning and public transports – it makes sense for the regional government to have more responsibility for reasons of proximity or local knowledge and a better match of policies with functional areas. For others, national governments may be in the best position for reasons of scale or capacity, to provide services efficiently3

● Lack of regional development strategies – strategies are either too fragmented or only top-down from the national government

● Need for improved spatial planning at the regional scale

● Need for improved synergies across sectoral policies (e.g. infrastructure, innovation, higher education, housing, labour market)

● Need for increased co-operation: to respond to the widening functional labour market region trend

● Local labour markets are sometimes smaller than individual municipalities, co-ordination challenges could be addressed by inter-municipal co-ordination arrangements in some cases

● Risk of insufficient transfer of resources and lack of regional capacities to conduct appropriate strategic planning in remote regions

Enhance local democracy

● Enhanced local democracy and transparency in decision making through direct election of regional bodies

● Limits of inter-municipal co-operation: horizontal co-operation comes at the cost of a less transparent decision-making process, as regional co-operation councils are indirectly elected, thus reducing voter control and transparency of the decision-making process

● No evidence of citizen support to regionalisation reforms due to the lack of consultation and the fact that counties are very old territorial units, with a strong identity

● More authority for regional government hence regional government becomes more important, yet turnout in regional elections is low. This challenge is more important for metropolitan and urban government which often have indirectly elected assemblies, making the “chain of accountability” between voters and representatives more indirect

Responsibilities assigned to regions

Compared to regions in unitary countries, those in federal countries tend to have greater policy scope and some authority to implement or shape major social policy areas, such as health, education, and social spending. In federal countries, regional tiers are generally in charge of borrowing, while in unitary countries there is greater variation. In contrast to local governments, which are often general-purpose, it is not uncommon for regions to be deconcentrated, possess dual executives, or have more limited autonomy, particularly in unitary countries.

Almost all regions that have general-purpose government play a role in the co-ordination and administration of education (often secondary), healthcare (specialised and hospital), social services, infrastructure and economic development, yet levels of authority vary widely. Some self-governing regions have very little policy scope, such as the Danish Regioner or Spanish Provincias. At the opposite end of the spectrum are the Australian States, Quebec in Canada, and the Swiss Cantons, which have full policy autonomy and also authority over immigration4, citizenship and residency.

Some regional governments depend completely on the centre and/or their municipalities for funding, while others have greater tax autonomy and thus extensive own resources, as well as constitutionally protected shares of central government transfers or taxes (tax sharing). Those with full control of major taxes include for example Canada, the Swiss cantons and the United States.

In the two-tier system of subnational governments, the regional level between the municipalities and the central government, because it operates on a larger scale usually provides services of regional interest, which benefit from economies of scale, generate spillovers, involve redistribution and are required to meet the same standards across the jurisdiction (OECD, 2017[21]). The regional tier may also facilitate co-operation and strategic planning.

In a three-tier system, as in France, Italy, Poland and Spain, the breakdown can be complex, sometimes resulting in duplication, overlap and co-ordination challenges. However, in recent decades, the intermediate level in a majority of these countries has lost many of its powers and responsibilities in favour of regions, which have gained more importance. Intermediate governments are now mainly responsible for administrative and delegated tasks, and have small budgets and in general no, or only limited, taxing powers.

Policy and governance implications of increasing regional authority

Regionalisation trends are clear in OECD countries and around the world, and they raise several policy and governance implications. They increase the need for co-ordination across government tiers and the need for clarification in the assignment of responsibilities, in order to avoid overlap. The fact that fiscal autonomy and control have not been strengthened at the same pace as other dimensions (like policy scope) implies that regions depend quite heavily on the national government for financing. This may curb the autonomy of regional government to the extent that central government transfers come with strings attached. It is quite likely that regionalisation trends will continue to progress in the coming decades, to generate economies of scale in public service delivery and provide more effective regional development policies, given persistent territorial disparities. However, in the current context of crisis of democracies, there is also a call for stronger municipalities, which is the government tier closest to citizens. Scale issues thus need to be balanced with accountability and democratic issues. One trend that might increase as well is the differentiation of responsibilities assigned to different regions. The following section will explore this trend in depth.

Growing asymmetric decentralisation

Asymmetric arrangements have been common since at least the 1950s and are still growing in popularity. In 1950, around half of the countries covered by the Regional Authority Index (RAI) had some kind of differentiated governance at the regional level. In 2010, almost two-thirds of the countries in the RAI had implemented asymmetric arrangements in some form. Asymmetric decentralisation, however, is in transition: between the 1950s and the 1970s asymmetric arrangements occurred mostly at a regional level, whereas the present trend seems to apply asymmetric decentralisation to major urban areas. There can be political, economic or administrative motives for asymmetric decentralisation (Bird and Ebel, 2006[22]). This section will review recent trends in asymmetric decentralisation in OECD countries.

What is asymmetric decentralisation?

Asymmetric decentralisation occurs when governments at the same subnational government level have different political, administrative or fiscal powers (Congleton, 2015[23]) (Figure 3.12). Political asymmetric decentralisation refers to situations where some regions or subnational governments have been given political self-rule that deviates from the norm or average assignment.

Figure 3.12. Three main types of asymmetric decentralisation

One common way to categorise asymmetric decentralisation is to divide the policies into de jure and de facto arrangements (Martinez-Vazquez, 2007[24]; Bird and Ebel, 2006[22]). De jure asymmetric decentralisation is based on the special legal status of a certain region. In some cases, the status is outlined in the constitution, but more often asymmetric treatment is established in the ordinary law (sometimes both).5 The regions with special de jure status often enjoy considerably broader political autonomy than other regions. Politically motivated asymmetry usually leads to administrative and fiscal asymmetry as well.

Even if subnational governments belonging to the same government tier are treated symmetrically in terms of the politico-legal system, there might still be de facto asymmetry in implementation. This administrative asymmetry often aims to take the different capacities of subnational governments into account. Administrative asymmetry may, for example, include sequencing a national policy so that the subnational governments that fulfil certain predetermined standards are given greater autonomy in spending and revenue. The rest of the subnational governments could then “grow into this role” over time.

Asymmetric fiscal arrangements consist of a wide variety of measures including special spending responsibilities, revenue bases or taxation rights and additional transfers. The main forms of asymmetric fiscal decentralisation can be summarised as follows:

  • Differential spending assignments. For example, some regions or subnational governments may be assigned tasks in specific services, which are otherwise provided by the central government or a higher level of subnational government.

  • Differential revenue autonomies. The subnational governments with more capacity may be given more tax autonomy than others. Asymmetric autonomy could also be used for collecting user fees or selling property.

  • Differential treatment in the transfer system. In this case, the regions with unique service needs or an exceptional operating environment may justify the use of special purpose grants or use of certain criteria in formula-based grants. Specific transfers may be used as an alternative to differential revenue autonomies.

  • Differential fiscal rules. Some subnational governments may be given more room for manoeuvre in borrowing for example. This could be the case if the subnational government has special needs for public investments and if it is capable of fulfilling its obligations.

An increasing trend in asymmetric decentralisation

During the last seven decades or so, asymmetric arrangements have become more common especially among unitary countries. In 1950, some 45% of the countries covered by the RAI and with regions had some kind of differentiated governance (autonomy, asymmetry or dependence). By 2010, this figure had increased to 62% (Hooghe et al., 2016[2]) (Figure 3.13).

Figure 3.13. Number of special autonomous regions, dependencies and asymmetric regions in 81 countries since 1950

Note: Asymmetric and special autonomous regions and dependencies are subject to a different kind of autonomy regime than standard regions. Dependencies are directly ruled by central government and have very limited autonomy. The decline in the number of dependencies is largely based on the change of dependencies into standard regions in South America.

Source: Hooghe, L. et al. (2016[2]), Measuring Regional Authority: A Postfunctionalist Theory of Governance, Oxford University Press, Oxford.

Asymmetric arrangements are increasing in metropolitan areas

Asymmetric decentralisation is often applied at three different scales: regional (state/province), metropolitan and local (Table 3.5). Asymmetric decentralisation trends are, however, changing: whereas during the past decades the asymmetric arrangements occurred mostly at the regional level, the present trend seems to apply asymmetric decentralisation to large cities or for selected local governments. One reason for this shift may be linked to a better understanding and acceptance of the benefits of urbanisation and agglomeration economies.

Table 3.5. Different scales of asymmetric decentralisation




● The most common form

● In 1950, around half of the countries covered by the Regional Authority Index (RAI) had some kind of differentiated governance at the regional level

● In 2010, almost two- thirds of countries in RAI had implemented asymmetric arrangements in some form

● Rising trend

● Long-term trend

● Since the 1950s: specific status for capital cities

● Since 1990s: increase in metropolitan governance to address specific challenges and capacities of metro regions

● 87 new metro structures created since 1991 compared to 14 between 1971 and 1991

● Different sets of responsibilities for different municipalities, depending on their capacities

● Municipal classifications, based on population, access to public services, budget, performance

● Urban/rural municipalities: classification may just statistical or lead to differentiation in funding or responsibilities

Asymmetric arrangements are increasing especially in unitary countries

Asymmetry is often a basic characteristic in federal countries. There are however different gradients of asymmetry: some federal countries are highly asymmetric, such as Canada, India, Russia and Spain. Some federations tend to favour symmetry, like Australia, Austria, Germany, Switzerland and the United States. However, even the most “symmetric federations” (e.g. Switzerland and the US) have elements of asymmetry (Bird, 2003[25]; Dafflon, 2006[26]).

In unitary states, symmetry is often one of the basic principles of the state, motivated by equity and integration of different parts of the country, such as in Chile and France. However, some unitary states have strong elements of asymmetry, in particular recognising a different status in territories with a strong history/identity (e.g. in Italy and the United Kingdom), as well as peripheral territories such as outermost regions, islands and outlying regions (e.g. Finland, France and Portugal).

Asymmetric decentralisation is increasing in unitary countries, based on new motives, including metropolitan governance arrangements, asymmetric administrative decentralisation and to give more responsibilities to regions with greater capacities. This is revealing a convergence between unitary and federal countries in the trend toward greater differentiated governance at the subnational level (Table 3.6).

Table 3.6. Asymmetric decentralisation/federalism

Federal countries

Unitary countries

● Asymmetry is often a basic characteristic of federations. There are however different degrees of asymmetry:

- Asymmetric: Belgium, Canada, India, Russia, Spain.

- More symmetric federations: Australia, Austria, Germany, Switzerland, United States.

● However, even the most “symmetric federations” have elements of asymmetry, e.g. United States.

● Symmetry is often one of the leading principles of a unitary state (with unity and integration).

● However, some unitary states have strong elements of asymmetry, in particular to recognising a different status in territories with a strong history/identity (Italy, United Kingdom) as well as peripheral territories such as outermost regions, islands, outlying regions (Finland, France, Portugal).

● Asymmetric decentralisation is increasing, based on new motivations.

Greater convergence in recent years

Examples of asymmetric decentralisation

Political asymmetric decentralisation

Political asymmetric decentralisation is commonly practised among OECD countries and elsewhere. This type of decentralisation generally takes place at the regional (state/province) level. In Italy, there are currently 5 regions (out of 20) with a special constitutional status, approved by the Italian parliament. Defined by the Italian constitution in 1948, the islands of Sardinia and Sicily and the Alpine regions of Friuli-Venezia Giulia,6 Trentino Alto Adige and Valle d’Aosta have been given special statute with the aim of avoiding separatist movements and to protect linguistic minorities. Asymmetric decentralisation means that these regions have broad legislative powers and considerable financial autonomy. For example, the Valle d’Aosta retains 90% of all its taxes, and Sicily at times 100%. While the 2001 reform of the Italian constitution gave all the Italian regions significant powers in the legislative field, the latest developments in Italy (especially since 2009) indicate that the decentralisation trend has reversed and that recentralisation seems to be a current trend (memorandum of Council of Europe 2015). If the centralisation trend prevails in all the regions except those with asymmetric treatment, the differences in self-government between the five special regions and the 15 ordinary regions are expected to increase. Moreover, some recent developments indicate that differentiation is still planned in some cases. The regions of Emilia Romagna, Lombardy and Veneto have recently signed a tentative deal with the Italian government for more autonomy, though this arrangement still needs to be accepted in Italy’s parliament before taking effect.7

The French territory of Corsica, previously one of the French departments, gained a special regional status in 1991. Corsica has its own institutions (the Corsican Assembly and the Executive Council of Corsica, each with a president) and more powers than other French regions. However, in general, the legislation governing the French regions applies also to Corsica if not defined otherwise in legislation. The regional reform implemented in 2015 reduced the number of French regions from 22 to 13, not including Corsica, however (Congress of Local and Regional Authorities 2016).

In Canada, while the Canadian constitution is based on a unified approach, it does enable asymmetric arrangements for Canadian provinces. More specifically, asymmetric decentralisation in Canada is mostly based on “menu federalism”, where the “opt in” or “opt out” choices are made available to all provinces. The province of Quebec has been using this option more frequently than other provinces (Smith 2005; Milne 2005). In general, Quebec is an example of de jure political asymmetric decentralisation that is based on historical, cultural and linguistic reasons. Being the largest province and second most populous of Canada’s ten provinces, French-speaking Quebec is an influential member of Canada’s federal government. This is in contrast to a situation in many other countries, where the asymmetric arrangements are applied mostly to regions of small economic importance. Quebec has had and used specific powers for example in healthcare provision, the pension system, with the position of the French language in government, and immigration screening. Although the secessionist movement in Quebec has gained more support since the 1980s, the majority of voters in referendums (in 1980 and 1995) for independence, voted against secession. As for the effectiveness of Canada’s asymmetric decentralisation model, while there seems to be continuous discussion about equal treatment of provinces in the federation, the main goal to keep the Canadian federation united has been reached (Bird and Vaillancourt, 2007[27]).

In the United Kingdom, since the devolution in 1998-99 of certain powers and responsibilities to regional elected bodies, local government organisation and functions are defined and reformed by the UK government (and parliament) for England, and by devolved nations for Northern Ireland, Scotland and Wales.

In Portugal, there is also an asymmetric organisation with two autonomous regions –Azores and Madeira – having legislative responsibilities as overseas territories (there is no self-governing region on the mainland). These autonomous regions are responsible for the financing and general supervision of local authorities within their territory, and also have the legislative power to create, dissolve and alter local government boundaries in accordance with the national laws.

Administrative asymmetric decentralisation

Sweden is an example of a highly decentralised country where the subnational government levels have important tasks and a strong autonomous position, and where an asymmetric and innovative approach to decentralisation has prevailed. In Sweden, there is a long history of asymmetric decentralisation. The geographic, demographic and socio-economic differences between counties are significant, and this has motivated the central government to permit bottom-up initiatives and reforms that aim to adjust governance structures and competencies according to territorial capacity. While an “across the board” regional reform is politically difficult, counties themselves have gradually and voluntarily implemented regional reform themselves.

In Finland, the government is currently planning a major regional reform to be launched in 2020.8 The reform would transfer health and social services, as well as regional development and labour services, from municipalities (healthcare and social services) and the central government (regional development and labour services) to the 18 new counties. Healthcare would be organised in an asymmetric fashion, as the most demanding hospital services, including emergency services, would be provided by only 12 regions. The remaining six regions were considered too weak to be able to arrange the most challenging specialised healthcare services. In addition, the Helsinki Metropolitan Area in the Uusimaa region would have a special arrangement in regional development and labour services.

In France, in December 2017, several French deputies presented a bill aimed at implementing a differentiation and simplification of standards/norms applicable in the territories. The purpose of the bill is to replace regulatory standards with measures adapted to a diversity of local situations. The proposed law is based on the observation that there is a proliferation of standards applicable to territories. The bill in question also proposes the creation of a principle of subsidiarity by entrusting the local authorities with the adaptation of the norms of application of the law. The bill proposes to allow differentiation of norms and standards (e.g. building, public works, environment, etc.) according to spatial or local government particularities.

In the United Kingdom, the “Cities and Local Government Devolution Act 2016” is considered an important step towards decentralisation. It makes various amendments to the 2009 Act to allow greater devolution of powers to combined authorities (housing, transport, planning and policing powers) and to introduce directly-elected mayors, thanks to “Devolution Deals”, which built on previous “city deals” (OECD, 2017[11]). The New Devolution Deals also include fiscal policies, which are discussed below in the section describing fiscal asymmetric decentralisation.

Another interesting but perhaps somewhat less common motivation for asymmetric decentralisation has been to deregulate and simplify government guidance of local governments. Over time, the accumulating normative regulation has become an obstacle for efficient public service provision and reforms. Deregulation of subnational government services is a difficult task, however. Therefore, many governments have often decided to proceed carefully, for example by experimenting. An example of this is Denmark, where between 2012 and 2015, nine local municipalities in Denmark were granted exemptions from government rules and documentation requirements in order to test new ways of carrying out their tasks, in a policy experiment known as the “Free Municipality” initiative.9

Fiscal asymmetric decentralisation

Over the past decades, the central government in Spain has devolved competencies asymmetrically to subnational governments. The so-called Foral Regime, autonomous communities (i.e. the Basque Country and Navarra), have a special constitutional status and an autonomous taxing authority, whereas other regions have limited local taxing authority. The Foral Regime regions are responsible for tax administration and have autonomy to set rates and bases (albeit with some limitations). The main tax bases such as income, corporate, wealth, inheritance and wealth transfers are fully administered by these regional governments. To compensate for the services that the central government provides these regions, the regions pay an amount to the central government. In terms of spending responsibilities, the Foral Regime does not differ from other Spanish regions (Garcia-Milà and McGuire, 2007[28]).

In the United Kingdom, the capital financing of local governments provides an example of asymmetric fiscal decentralisation. While borrowing from the Public Works Loan Board has been the most common form of capital financing of local governments in the United Kingdom, new ways to access capital finance for local infrastructure investment have also been developed. For instance, in 2013-14, the UK government has introduced tax increment financing schemes, which allow local authorities to borrow against future growth in business rate receipts. Another example is the “New Development Deal”, which benefits particularly cities with special status as defined by central government. The 2012 Budget set a limit of GBP 150 million that could be borrowed via New Development Deals: the funding would only be available to “core cities” (House of Commons, 2016).

In South Africa, the 278 municipalities have different fiscal and administrative capacities. The South African government has adopted a differentiated decentralisation approach, particularly with respect to municipal funding. In order to accomplish this, several classifications are used to group municipalities. The 1996 constitution provides for three categories: A (metropolitan municipalities), B (local municipalities) and C (district municipalities). Other classifications exist in other areas, such as the Municipal Infrastructure Investment Framework Investment that defines seven categories based on spatial characteristics, size of institution and budget, and population variables, among others. The National Treasury also classifies municipalities into six “performance groups” using economic, demographic and performance variables such as access to basic services, poverty rate, municipal viability, staff vacancy, municipal debt, population density and size of the municipality’s economy. As underlined by the Financial and Fiscal Commission, the different classification methodologies highlighted above recognise that municipalities need differentiated approaches that take into consideration their different characteristics and needs. However, the South African Financial and Fiscal Commission has argued that a differentiation approach is not always clear, and some classifications are not always useful for making decisions or allocating resources. This is because categorisation often depends on the policy purpose and often detracts from looking at the links between rural and urban municipalities. In addition, the large volume of classification methodologies may undermine co-ordinated decision-making and intervention strategies (OECD, 2017[21]).

In Colombia, the country’s 1 101 municipalities are responsible for providing electricity, urban transport, cadastre (land registry), local planning and municipal police. Municipalities are grouped into “certified” or “non-certified” units: only certified municipalities are allowed to provide important services such as health, education, water and sanitation. If the municipality is considered too weak or it is otherwise ineligible for service provision, the service is provided by the upper (department) government level (OECD-UCLG, 2016[20]; OECD, 2014[29]).

In Norway, the small rural local governments with substantial tax revenue from hydropower plants provide another example of asymmetric fiscal decentralisation. The municipalities where the power plants are located receive taxes and other revenues from the power company. These revenues are considered compensation for environmental damages and have been important for generating local support for projects that are profitable for society as a whole. The average total revenue per capita among local governments with hydropower revenue was NOK 32 600 (USD 6 520) in 2007. In comparison, the revenues for all other local governments were NOK 28 300 (USD 5430) (Borge, Parmer and Torvik, 2015[30]).

In Sweden, the intergovernmental grant system is an example of “Robin Hood” equalisation, which evens out a considerable share of differences in subnational government revenue bases and service costs. A rather detailed formula is used to define the grants for subnational governments. While the transfer system is primarily based on general grants, also some discretionary and earmarked grants have been used. The so-called “structural grants” are related to regional policy and their aim is to strengthen municipalities with a small population, with decreased population and/or with a problematic labour market. Structural grants mainly benefit the more remote municipalities, for example in Norrland County (OECD, 2017[31]).

Table 3.7. Asymmetric decentralisation by type and scale, some examples of practices





Italy: Five regions with special constitutional status.

France: The French territory of Corsica has a special regional status.

Portugal: The autonomous regions of Madeira and Azores have special legislative responsibilities

Canada: Province of Quebec has used actively the “opt in, opt out” choices available to all provinces.

Sweden: A voluntary regionalisation reform from 1990s until 2018 in terms of political representation and responsibilities in different regions and in different phases. Since the beginning of 2019 all Swedish counties have been responsible (by law) for regional development.

Spain: The autonomous taxing authority of “Foral Regime” autonomous communities.



France: 14 metropolises will be granted greater responsibilities than “standard” municipalities

Italy: 14 metropolitan cities created to administer large urban areas.

UK: Special investment funding for core cities.

US: Some states assign a portion of state tax revenues to municipalities with a substantial share of the state population (New York City, St. Louis, Kansas City).

Germany: The “City-States” like Berlin have both state and local government responsibilities and revenues.



Denmark: Free municipality experiment in order to simplify regulation.

Czech Republic: Limited number of municipalities perform central government delegated functions on behalf of smaller surrounding municipalities.

Colombia: Royalty revenues for municipalities in certain mining/oil regions.

Norway: Hydropower revenues for specific municipalities.

Sweden: Special grants for the most rural/remote municipalities.

Source: Allain-Dupré, D., I. Chatry and A. Moisio (forthcoming[32]), “Asymmetric decentralisation: Trends, challenges and policy implications”, OECD, Paris.

Policy implications

There are clear trends showing greater use of asymmetric decentralisation arrangements in the OECD and around the world. The trends also indicate that once adopted, asymmetric arrangements are maintained in the long term. Of the countries with differentiated decentralisation arrangements in 1950, every one still had differentiated regions in 2010, while 10 of the 24 countries without differentiation in 1950 included differentiated regions by 2010 (Hooghe et al., 2016[2]).

There is no blueprint or optimal strategy for choosing between an asymmetric or symmetric approach because the optimal strategy is usually case specific and depends on local circumstances. Nonetheless, some observations and conclusions can be made based on the economics literature and experiences from practical implementation. This section aims to shed light on the pros and cons of asymmetric arrangements based on key policy considerations. There are benefits and challenges associated with asymmetric decentralisation. Benefits are linked to the fact that institutional and fiscal frameworks are better targeted to local capacities, and may allow a better response to local needs. In general, asymmetric decentralisation favours experimentation, learning-by-doing and innovation in policymaking. Ultimately, it represents an advanced form of place-based policy. The challenges of asymmetric decentralisation are associated with the cost of co-ordinating a complex system that may be unclear for citizens and lead to accountability challenges, notably at the metropolitan level. Other potential challenges include the increasing disparities in capacity across regions, which would call for adequate equalisation systems.

Sequencing decentralisation reforms may help central and subnational governments learn from successes and mistakes and take steps to revise the reform if needed. In this way asymmetric decentralisation can be seen as a form of experimentation and “menu federalism”, where subnational governments voluntarily choose the tasks that best serve their own interests (Congleton, 2015[23]). Asymmetric decentralisation also helps implement tailored governance frameworks and place-based regional policies. For example, the effects of major exogenous shocks such as natural disasters or climate change usually affect different regions differently. Accommodating diverse preferences for political and fiscal autonomy across regions may also mitigate separatist movements and help maintain political stability.

Perhaps the most significant risk of asymmetric decentralisation relates to the fact that, by definition, asymmetric arrangements do not directly promote equal treatment of subnational governments and citizens. In some cases, asymmetric decentralisation may result in a perception that asymmetry means deviation from an overall objective of equality. Spending assignments with a clear redistributive function such as education, health and social services could be examples of such services. In addition, there may be a risk that asymmetric decentralisation is perceived as support for the wealthiest regions or subnational governments. The “favoured” governments may be able to attract citizens and business from other subnational governments, which could accelerate the differentiated economic and social development between regions (Congleton, 2006[33]). If widely applied, asymmetric arrangements may reduce the transparency and accountability of governance and result in complex administrative systems (Bahl and Martinez-Vazquez, 2005[34]). As in with decentralisation in general, the effects and outcomes of asymmetric decentralisation depend on implementation.

Shared responsibilities, mutual dependence and a renewed role for central governments

Decentralisation policies affect subnational governments, certainly, but also national governments. They imply a renewed role for central governments, one that is more strategic but also more focused on setting the conditions for proper co-ordination and alignment of policy objectives, monitoring the performance of regions and cities, and ensuring the balanced development of all parts of the national territory. Given that most responsibilities are shared, decentralisation policies imply managing mutual dependence to achieve common objectives.

Most responsibilities are being shared between the central and subnational governments

As highlighted in Chapter 2, most responsibilities – apart from defence and monetary policy – are shared among levels of government. The extent of responsibility sharing also depends on the service in question. For example, responsibilities tend to be more frequently shared in public transport than in childcare or elderly care (OECD, 2016[35]). Some examples of commonly shared responsibilities include:

  • Physical infrastructure and its related public investment are among the most commonly shared responsibilities (OECD, 2014[36]). Subnational governments are generally responsible for local roads and local transport infrastructure. Higher levels of government generally manage investments with high externalities.

  • Education is the most common major public service to be shared substantively across multiple tiers of government (OECD, 2016[35]; OECD, 2016[37]). In most OECD countries, lower levels of government are responsible for managing and funding lower levels of schooling (mainly pre-elementary, primary and sometimes lower secondary education) whereas responsibility for secondary, and in particular upper secondary, schooling rests more often with provincial/regional or central levels. In addition, central governments establish framework legislation while local governments are usually responsible for maintaining and constructing physical infrastructure and the paying staff, though in many cases without actual authority over hiring or salary.

  • Healthcare is usually more centralised than education because of high system complexity, with specific roles for different levels of government but often less autonomy (OECD, 2016[37]).

  • Social assistance or welfare is much more varied and is more likely to be jointly provided because of the strongly redistributive character of the service.

  • Land use planning is predominantly a local task, even though several countries use guiding land use plans prepared at the inter-municipal or regional levels. National and regional governments both focus primarily on strategic planning and the provision of policy guidelines. Most countries do not prepare land use plans for their entire territory but are more likely to focus on land use plans for areas of particular importance (OECD, 2017[38]).

Table 3.8. Shared responsibilities across levels of government in federal and unitary countries
Proportion of decisions where more than one government level is involved (%)



Long-term care

Transport services

Social housing








Russian Federation





South Africa



















































Average across federal countries












New Zealand










































South Korea






















Average across unitary countries






Note: The above calculations account for any “not applicable” or unanswered responses for each country.

Source: OECD (forthcoming[12]), The Spending Power of Subnational Governments across Five Policy Sectors, OECD Publishing, Paris.

Decentralisation implies a renewed role for central governments

The impact of decentralisation on the central government is often underestimated. Decentralisation represents substantial challenges for central governments as it requires profound changes of organisation, practices, culture and skills within the central government itself (Devas and Delay, 2006[39]).

Not only do decentralisation reforms affect central governments in the capital, but they also affect central government representation at the territorial level as well.

Failing to take the full measure of this may be detrimental to reform success, slowing down or modifying the reform process (OECD, 2017[11]). The fiercest opponent to decentralisation efforts is often the central government itself, as decentralisation may be perceived as a threat (Box 3.5).

Decentralisation reforms imply a more strategic role for national governments, which focus on putting in place the strategies, framework conditions and regulations, incentives – rather than in the implementation of policies. It involves a shift from a direct role in service delivery to one of enabling, advising and assisting, ensuring consistency, facilitating the work of local governments and sometimes helping share good practices across local governments. This requires building new capacity at the central government level, able to cope with these new functions, which cover a large area of sectors.

Box 3.5. Resistance from central government to decentralisation reforms

Resistance to decentralisation reform is often strong from central government itself, even if it is the main promoter of the reform and in charge of its design and implementation. Departments in charge of leading decentralisation processes may be unsuccessful at persuading other departments or ministers to relinquish powers.

There are different reasons for this. First, national government tends to have limited trust in subnational government competency and accountability for failure (Gash, Randall and Sims, 2014[40]). Second, decentralisation may represent a loss in terms of public employees and control on public affairs. Third, decentralisation requires profound changes in organisation, practices, culture and skills within the central government itself.

Decentralisation implies a reduction in central government employees. It may also directly affect civil servant jobs, particularly if they are transferred to subnational governments, which could involve a less advantageous shift in status, wages, pensions and other benefits. In Ukraine for example, it is expected that the decentralisation reform will reduce the number of central government civil servants by about 30% as numerous functions will be transferred to subnational levels (OECD, 2018[41]). In Chile, the creation of self-governing regions, with the democratic and direct election of regional governors will also have a strong impact on the regional representation offices of national ministries (SEREMI) and central government regional agencies (OECD, 2017[21]).

In France, this issue has only recently been addressed. Decentralisation initiatives have not been accompanied by a reform of central government structure, especially at the local level. This situation has generated duplications in responsibilities, services and staff between the central and deconcentrated, subnational units (OECD, 2017[11]).

Sources: Gash, T., J. Randall and S. Sims (2014[40]), Achieving Political Decentralisation: Lessons from 30 years of Attempting to Devolve Political Power in the UK, Institute for Government; OECD (2018[41]), Maintaining the Momentum of Decentralisation in Ukraine,; OECD (2017), Making Decentralisation Work in Chile: Towards Stronger Municipalities,; OECD (2017), Multi-level Governance Reforms: Overview of OECD Country Experiences,

National/federal governments play a crucial role in ensuring that decentralisation does not widen disparities in terms of access and quality to the public services provided by subnational governments in their new functions. There is a risk exists of significant heterogeneity in financial and human capacities given differences in local capacities and performance among subnational governments. In a context of reinforced decentralisation, central governments must address this challenge with appropriate instruments aimed at monitoring, diagnosing and improving the performance of local services. This could include such instruments as management and certification models based on minimum quality standards and performance indicators for basic local services. In Chile, for example, a key factor in the reform of the municipal management system is setting minimum standards for municipal services (Servicios Municipales Garantizados, SEMUG) to reduce horizontal disparities among municipalities. Initially, the SEMUG is composed of seven municipal services called “the first generation of guaranteed minimum services”. These services represent a high impact on the community, or high costs or income for the municipality. The minimum standards encompassed a basic level of quantity and quality for a common set of services to be guaranteed by all Chilean municipalities, and accessible to all citizens regardless of where in the country they choose to live. The 7 selected services include 22 standards and 47 indicators (OECD, 2017[21]). In Norway, the KOSTRA performance measurement system electronically publishes result within a month of receipt from the municipalities, in this way supporting municipal public service provision and capacity. Some other OECD countries such as Italy, the Netherlands and the United States have gone beyond the definition of minimum standards by developing service charters to promote the continuous improvement of public services and to make clear commitments to the quality standards that citizens can expect.

Another challenge for the central government is linked to human resources management, to make sure that decentralisation does not generate a two-speed civil service. In Chile, for example, although municipal governments now have the same public sector employment arrangements as the central government, they do not have the same employment conditions. Consequently, municipal capacity remains weak as attracting and retaining a highly-skilled workforce is challenging. Low salaries, limited career opportunities and politics are just some of the factors that hinder the attractiveness of municipal public administrations as employers, and thus represent a challenge for the success of decentralisation (OECD, 2017[21]). Decentralisation should involve a convergence of civil service status, salaries and advantages at central and subnational levels. Disparities between the central government and subnational sector may be significant. Central governments have a crucial regulatory and co-ordination role to play in ensuring that subnational governments have the necessary autonomy and flexibility in this area. This can include the ability: to define the general framework for public employment (i.e. the legal, financial and social conditions of the employees, procedures of recruitment, staff categories and salary scales, retirement conditions, ethics rules, workforce planning, performance systems, etc.); to develop training and learning programmes to bridge capacity gaps; to favour staff mobility across levels of government, etc.

Box 3.6. OECD country experience in central-local co-ordination for pay-setting at the subnational government level

The OECD has identified several reasons why a national government may seek to influence or control remuneration and other employment conditions for staff in subnational administrations. Differences in employment conditions might: i) hamper mobility across public administrations and government levels; ii) as a means to limit or cap the growth of public expenditure; or iii) in order to ensure coherence in public employment conditions. Countries are using a number of instruments to oversee municipal payrolls while at the same time granting subnational governments flexibility to set their own wages.

  • In Spain, the law that regulates the civil service also regulates the structure of the pay system for civil servants at the national and subnational levels. The annual remuneration increase is contained in the general state budget. Contracted employees are employed under normal labour market conditions.

  • In France, each local government can determine remuneration and other employment conditions for its employees, but their actions are regulated by law and by the fairly complex regulations for the French corps (or career) system.

  • Subnational governments in Denmark set their own wages. However, the State Employers Authority maintains an informal, ongoing dialogue with municipal and regional associations that function as central employers for the subnational administration. The State Employers Authority is also represented on the municipal and regional Boards of Wages and Tariffs that function as employer representatives in negotiations with the unions in these sectors and has veto power on the regional board.

Sources: OECD (2008[42]), Challenges of Human Resource Management for Multi-level Government; OECD (2017[21]), Making Decentralisation Work in Chile: Towards Stronger Municipalities,

The central government also needs to renew its supervisory and monitoring role vis-à-vis subnational governments, especially in legal and fiscal matters. Control should be based on posteriori rather than a priori control. As far as budget and financial supervision and control are concerned, a posteriori controls are essential in a context of increased fiscal decentralisation and greater autonomy. The central government also has a role to play in accordance with the principle of local autonomy. Financial audits are necessary to assess the quality of financial reporting and the reliability and accuracy of financial information and management provided by subnational governments. But with respect to external audits, state financial supervision and the control system over subnational governments should be adapted to the new decentralisation context, for example in liaison with an external and independent audit institution. In several OECD countries, the supreme audit institution can audit both state and local government budgets on the expenditure and revenue side. This is the case in France (Cour des Comptes), Germany, Italy (Corte dei Conti), Poland (NIK) and Portugal, for example. Some national supreme audit chambers have a network of regional chambers as well (e.g. the Chambres régionales des comptes in France) (OECD, 2018[41]; 2017[21]).

Box 3.7. Central government’s evolving role in the context of decentralisation

France has started to take into account this need through the 2010 reform of the state (REATE) which aimed at streamlining the territorial administration and at improving the old and complex relationships between deconcentration and decentralisation. In 2015, the French regional reform was also an opportunity to think about the modernisation of the functioning of government services at the regional level and to redefine their role and missions (Charter for Deconcentration). More recently, a new impulse has been given in this direction by the new French government, based on the recommendations of the “Public Action Committee 2022” (CAP 22), concerned with the need to draw all the consequences of decentralisation, eliminate the duplication of competencies between the central and subnational governments, and rethink the role and the areas and methods of intervention of the state at regional and local levels.

Finally, the role of national governments to establish adequate co-ordination mechanisms – ones that support aligning objectives and policies across and among levels of government is fundamental in this changing context. The same is true for their ability to design instruments for effective intergovernmental relations, co-operation and dialogue, as these can also build trust in a reform process (Box 3.8).

This is particularly important for regional development policies. Responsibilities for regional development are now more fragmented among national, regional and local actors, which have various resources, agendas, and legal or political standing. In this context, the role of the central government is increasingly important for providing an overarching framework and guidelines for regional development policies. It also needs overseeing co-ordination mechanisms within which regional policy can be formulated and implemented.

Ensuring that objectives and priorities are aligned, particularly in areas that support regional development but where responsibilities and/or interests overlap (e.g. economic development, transport, health and education) requires vertical co-ordination mechanisms that foster a partnership-based relationship among levels of government.

Some OECD countries have addressed these concerns early on in their decentralisation processes, improving governance structures between levels of government as well as across sectors (see Chapter 5). In Denmark, the Regional Growth Forum integrates local, regional, national and EU development activities within a single, programme-based policy structure. Sweden’s Regional Growth Policy aims to improve local and regional competitiveness across all regions through regional programmes and enhanced regional and sectoral co-ordination (OECD, 2010[43]).

Box 3.8. The renewed role of the central government for managing regional development policies

The following central government roles have emerged under the new paradigm of regional development policies:

  • Facilitate consensus-building and coherence between regions and sectors including defining objectives, time frames and spatial horizons.

  • Gather and analyse appropriate data and information and co-ordinate discussions and databases concerning needs and opportunities: facilitating dialogue among policymakers.

  • Develop legal, fiscal and administrative frameworks: frameworks or “grand rules” which manage the complexity, plurality and tangled hierarchy characteristic of most modes of co-ordination.

  • Serve as a “court of appeal” for disputes among sectors and regions: including taking political responsibility for the final decision, especially in the event of a governance failure.

  • Seek to re-balance power differentials among sectors, regions and levels of governments: for the proper functioning of the overall governance system, the national government can and should help weaker entities establish capacity building strategies (including training provided by the central government).

  • Evaluate and monitor policy results: closing information gaps and improving the quality of decision-making by actors at all levels of government.

Source: OECD (2010[43])Regional Development Policies in OECD Countries

Ensuring balanced development in all parts of the national territory

In decentralised settings, national governments have a key role in ensuring a balanced development in all parts of the national territory and in minimising the potential risk of increased disparities (see Chapter 4). This can be achieved through active regional development policies at the national level and well-designed equalisation policies, which promote development efforts of subnational governments. Fiscal equalisation aims to correct imbalances between subnational governments, thus fostering equity between territories, be they regions or localities.

Fiscal disparities can be one of two kinds: differences in revenue-raising capacity and differences in public service expenditure needs. In the first category, unequal tax-raising capacities come from differences in per capita GDP across jurisdictions. In the second, they are related to specific geographical factors (mountains, islands, isolated or low-density areas, etc.) or special groups such as children, the elderly, the disabled, etc. that explain a higher cost per service unit, raising the overall cost of public services.

Equalisation mechanisms are widely used in OECD federal and unitary countries, either through vertical transfers (from the central/federal government to subnational governments that are financially weak) or through horizontal transfers (from the wealthiest subnational governments to the poorest), or both (see Chapter 5). Equalisation schemes also aim to develop national standards to ensure equal access to local services and a minimum level of quality in the provision of local public services.

Most OECD countries apply various equalisation arrangements, combining horizontal and vertical equalisation with tax revenue and cost equalisation. Vertical equalisation tends to prevail, though both systems have their advantages and disadvantages (OECD, 2013[44]). Tax revenue equalisation and cost equalisation systems are roughly the same size, although tax revenue disparities are between four and six times larger than cost disparities (Kim and Lotz, 2008[45]). One of the main difficulties in establishing an equalisation system is the way tax-raising capacity and/or the cost of services is measured (see Chapter 5).

  • In Germany, the principle of uniformity of living conditions for Germans throughout the Federation and equalisation is enshrined in the constitution. In Switzerland, the equalisation system was established in 1958 by the Federal Law on Financial Equalisation and thoroughly reformed in 2008.

  • In Australia, fiscal equalisation is not mentioned in the constitution but the objective of equalisation is strong. The country aims to ensure that all states have the same fiscal capacity per capita to provide the same services and infrastructure to all residents.

  • In Italy, the horizontal equalisation mechanism between the regions created in 2001 was reformed in the framework of Law 42/2009, strengthening vertical equalisation through a state-run fund. It aims to ensure the coverage of essential public services (e.g. health, education and social assistance) in regions with low tax revenues.

  • In Sweden, the vertical and horizontal equalisation system is based on five allocations, including income equalisation and cost equalisation. The system is regularly reviewed with an eye on assessing any counterproductive and dissuasive effects the equalisation system may have and correcting any drift, in one direction or the other.

In most instances, the effect of equalisation on the reduction of fiscal disparities between subnational governments is substantial across the OECD (Box 3.9). However, equalisation is frequently the subject of technical and political debate, and is often contested for its complexity, lack of transparency and its potentially negative incentive effects on tax base development and spending levels (OECD, 2013[44]; 2018[46]).

Box 3.9. Equalisation in the OECD countries: Some key facts

For a set of 15 OECD countries, equalisation amounts to around 2.3% of GDP, 4.8% of total government expenditure and around half of all intergovernmental grant fiscal equalisation transfers, confirming that there is ample room for improving equalisation.

Across OECD countries, equalisation has a strong redistributive effect: on average, fiscal equalisation diminishes disparities in revenue-raising capacity – as measured by the Gini coefficient or variation coefficient – by almost two-thirds, from 29% to 10%. In some countries – such as Australia, Germany and Sweden – revenue-raising disparities are virtually eliminated. After equalisation, fiscal disparities are clearly narrower than economic disparities, as measured by regional GDP. In other words, the ability to provide public services is more evenly distributed than economic output.

Source: OECD (2013[44])Fiscal Federalism 2014: Making Decentralisation Work


[17] Ahrend, R., C. Gamper and A. Schumann (2014), “The OECD Metropolitan Governance Survey: A Quantitative Description of Governance Structures in large Urban Agglomerations”, OECD Regional Development Working Papers, No. 2014/04, OECD Publishing, Paris,

[32] Allain-Dupré, D., I. Chatry and A. Moisio (forthcoming), “Asymmetric decentralisation: Trends, challenges and policy implications”, OECD, Paris.

[34] Bahl, R. and J. Martinez-Vazquez (2005), Sequencing Fiscal Decentralization.

[10] Bergvall, D. et al. (2006), “Intergovernmental Transfers and Decentralised Public Spending”, OECD Journal on Budgeting, Vol. 5/4,

[25] Bird, R. (2003), Asymmetric Fiscal Decentralization: Glue or Solvent?.

[22] Bird, R. and R. Ebel (2006), “Subsidiarity, solidarity and asymmetry: Aspects of the problem”, in Fiscal Fragmentation in Decentralised Countries, Edward Elgar Publishing Limited and World Bank.

[27] Bird, R. and F. Vaillancourt (2007), “Expenditure-based equalization transfers”, Fiscal Equalization,

[30] Borge, L., P. Parmer and R. Torvik (2015), “Local natural resource curse?”, Journal of Public Economics, Vol. 131/C, pp. 101-114.

[23] Congleton, R. (2015), “Asymmetric federalism: Constitutional and fiscal exchange and the political economy of decentralisation”, in Ahmad, E. and G. Brosio (eds.), Handbook of Multilevel Finance, Edward Elgar Publishing Limited.

[33] Congleton, R. (2006), “Asymmetric federalism and the political economy of decentralisation”, in Ehtisham, A. and G. Brosio (eds.), Handbook of Fiscal Federalism, Edward Elgar Publishing Limited.

[26] Dafflon, B. (2006), “The assignment of functions to decentralised government: from theory to practice”, in Bird, R. and R. Ebel (eds.), Fiscal Fragmentation in Decentralised Countries: Subsidiarity, Solidarity and Asymmetry.

[4] de Mello, L. (2018), “The global crisis and intergovernmental relations: Revisiting the centralisation-decentralisation debate ten years on”, Working Papers. Collection A: Public economics, governance and decentralization 1803, Universidade de Vigo, GEN Governance and Economics research Network..

[39] Devas, N. and S. Delay (2006), “Local democracy and the challenges of decentralising the state: An international perspective”, Local Government Studies, Vol. 32/5, pp. 677-695,

[28] Garcia-Milà, T. and T. McGuire (2007), “Fiscal decentralization in Spain: An asymmetric transition to democracy”, in Bird, R. and R. Ebel (eds.), Fiscal Fragmentation in Decentralized Countries: Subsidiarity, Solidarity and Asymmetry, Edward Elgar Publishing.

[40] Gash, T., J. Randall and S. Sims (2014), Achieving Political Decentralisation: Lessons from 30 years of Attempting to Devolve Political Power in the UK, Institute for Government.

[19] Hooghe, L., G. Marks and A. Schakel (2010), The Rise of Regional Authority. A Comparative Study of 42 Democracies, Routledge, London.

[2] Hooghe, L. et al. (2016), Measuring Regional Authority: A Postfunctionalist Theory of Governance, Oxford University Press, Oxford.

[1] Ivanyna, M. and A. Shah (2014), “How close is your government to its people? Worldwide indicators on localization and decentralization”, Economics,

[14] Keuffer, N. and A. Ladner (2018), The Local Autonomy Index Project – Extent, Patterns and Effects of Local Autonomy in Europe.

[45] Kim, J. and J. Lotz (2008), Measuring Local Government Expenditure Needs: The Copenhagen Workshop, Korea Institute of Public Finance and the Danish Ministry of Social Welfare.

[3] Ladner, A., N. Keuffer and H. Baldersheim (2016), “Measuring local autonomy in 39 countries (1990–2014)”, Regional and Federal Studies, Vol. 26/3, pp. 321-357,

[24] Martinez-Vazquez, J. (2007), Asymmetric Federalism in Russia: Cure or Poison?,

[6] OECD (2018), Key Data on Local and Regional Governments in the European Union, OECD and EU,

[41] OECD (2018), Maintaining the Momentum of Decentralisation in Ukraine, OECD Multi-level Governance Studies, OECD Publishing, Paris,

[46] OECD (2018), Practical Methodological Guide for the World Observatory on Subnational Government Finance and Investment, OECD, Paris.

[15] OECD (2018), Subnational Governments in OECD Countries: Key Data (brochure and database), OECD, Paris,

[38] OECD (2017), Governance of Land Use, OECD Publishing, Paris.

[21] OECD (2017), Making Decentralisation Work in Chile: Towards Stronger Municipalities, OECD Multi-level Governance Studies, OECD Publishing, Paris,

[11] OECD (2017), Multi-level Governance Reforms: Overview of OECD Country Experiences, OECD Multi-level Governance Studies, OECD Publishing, Paris,

[5] OECD (2017), OECD Territorial Reviews: Sweden 2017: Monitoring Progress in Multi-level Governance and Rural Policy, OECD Territorial Reviews, OECD Publishing, Paris,

[31] OECD (2017), OECD Territorial Reviews: Sweden 2017: Monitoring Progress in Multi-level Governance and Rural Policy, OECD Territorial Reviews, OECD Publishing, Paris,

[35] OECD (2016), Assigning Responsibilities across Government Levels, Note by the Secretariat COM/CTPA/ECO/GOV(2016)3.

[37] OECD (2016), OECD Regions at a Glance 2016, OECD Publishing, Paris,

[16] OECD (2015), Governing the City, OECD Publishing, Paris,

[36] OECD (2014), OECD Recommendation on Effective Public Investment across Levels of Government,

[29] OECD (2014), OECD Territorial Reviews: Colombia 2014, OECD Territorial Reviews, OECD Publishing, Paris,

[8] OECD (2014), OECD Territorial Reviews: Netherlands 2014, OECD Territorial Reviews, OECD Publishing, Paris,

[44] OECD (2013), Fiscal Federalism 2014: Making Decentralisation Work, OECD Publishing, Paris,

[43] OECD (2010), Regional Development Policies in OECD Countries, OECD Publishing, Paris,

[42] OECD (2008), Challenges of Human Resource Management for Multi-level Government.

[13] OECD (n.d.), OECD Fiscal Decentralisation Database,

[12] OECD (forthcoming), The Spending Power of Subnational Governments across Five Policy Sectors, OECD Publishing, Paris.

[9] OECD/KIPF (2016), Fiscal Federalism 2016: Making Decentralisation Work, OECD Publishing, Paris,

[20] OECD-UCLG (2016), Subnational Governments around the World: Structure and Finance, OECD, Paris,

[7] Quinn, B. (2015), “Local government reform: Plus ça change, plus c’est la même chose”, Administration, Vol. 63/2, pp. 7–29,

[18] Schakel, A. et al. (2018), Final Report on Updating the Regional Authority Index (RAI) for Forty-Five Countries (2010-2016).

[47] tovar (n.d.), .


← 1. An important exception is combined authorities in the United Kingdom. A reform in 2016 made it possible to establish direct elections for a mayor and combined authorities with directly elected mayors can increase business rates by two pence in the pound. This reform went into effect in six combined authorities when their mayors were elected on 4 May 2017. This falls outside the scope of the RAI update for 2010-16.

← 2. For the 48 countries in the RAI dataset for the full 60 years, regional authority increased from an average of 8.1 to 12.6.

← 3. In particular, there is a recent trend in OECD countries such as Denmark, Germany and Norway to recentralise healthcare provision. This option was also discussed in Sweden. The proposal was that healthcare might be reorganised, with primary care entrusted to municipalities and specialised care brought closer to research facilities. This would mean the recentralisation of specialised healthcare. However, Swedish regional reform is not moving in that direction, although some options have been discussed.

← 4. While Quebec does have substantial authorities over immigrant selection, the final decision to issue a visa remains a federal matter. Also, the Quebec government does not have authorities over Citizenship or Residency – these are both federal jurisdictions.

← 5. The division of asymmetric arrangements (with legal basis) into ones that are based on constitutional status and others that are based on ordinary law is of course a simplified description of reality. In many cases, such as Spain for example, the regions may have special status both by constitution and ordinary law.

← 6. The region of Friuli-Venezia Giulia has had a special status in since 1963.

← 7. Source: (accessed on 16 May 2018).

← 8. Finnish Ministry of Health and Social Affairs Internet pages: (accessed in February 2018).

← 9. In fact, the tradition of experimenting using the “free municipality” initiatives had already begun in several Nordic countries in the 1980s. The first free municipality trials were implemented in Sweden in 1984, Denmark in 1985, Norway in 1987 and Finland in 1989 (Swedish Government, 1991). In Norway, there has been a law for continuous experimenting since 1993. The law allows voluntary experimenting in municipalities, counties and central government.

End of the section – Back to iLibrary publication page